In current broadband networks a network provider offers a physical access connection from an end users home to the broadband network itself. Through this connection multiple services such as Internet, Voice over IP (Internet Protocol), TV over IP, telephony over IP and other services are offered.
Each service may require different termination equipment so called customer premises equipment (CPE) in the end users home. For example, to surf on the Internet a PC is utilized, and to watch TV a “setop box” is utilized.
In order to be able to connect all these devices to the same single physical connection a low cost customer premise switch (CPS) is utilized. This CPS is typically an Ethernet switch or ADSL modem or other device with Ethernet bridging capabilities.
The CPU power and functionality in the CPS switch is often limited. In addition, since the switch is physically located within the end user home, control of traffic, access control to the broadband network etc., for each and every CPE, is typically not a task for the CPS as the risk of fraud would increase. Therefore, broadband access equipment, such as an advanced router or switch, is placed at a central location close to the end user but in a location controlled by the broadband network operator. This device is responsible for creating and delivering the mentioned services to the end user, including access control such as does the end user subscribe to the service, bandwidth control, traffic prioritization and other similar tasks.
When facing multiple client/customer equipment such as a PC and a setop box, it is important for the broadband access equipment to be able to identify each CPE as each of them subscribe to a different service. The PC subscriber may not watch TV channels; the setop box subscriber may not surf on the Internet.
In current broadband networks, identification of the CPE is typically accomplished by using the MAC address (medium access control address, unique network card address) of the specific CPE, or parameters such as vendor id field in a DHCP request (dynamic host configuration protocol request) from the client equipment/CPE to obtain an IP address. When the client equipment requests an IP address, it is typically also time for the broadband access equipment to perform an access control to determine if the client equipment is permitted on the network, and if so which service parameters to install to control the traffic to that specific CPE.
There are several problems with the identification mechanism mentioned. For example, the MAC address is typically bound to a specific network card. If an end user changes PC or replaces the network card, a new MAC address is introduced, and this may require the end user to contact the customer service of the broadband network to have the MAC address record updated. Since the MAC address is a rather complex combination of digits and letters, and since the MAC address is difficult to obtain from within Windows® or to read on the outside of the PC chassis, this quickly becomes too technical for many everyday users.
Utilization of the vendor id field is another option, but as the field name implies, a vendor id from one vendor may not be the same as a vendor id from another vendor. In order to distinguish a PC from a setop box, or even a PC with Windows® from a PC with Linux, the broadband operator has virtually to understand all common vendor identifications (ids) from client equipment that the end user may purchase and use. Again, this is quite complex.
The patent Application US 20060036733 A1 to Fujimoto et al discloses, that a DHCP server can query a CPS regarding the MAC table, which can be utilized to determine if a specific CPE is allowed to connect to the network or not.